Pin-tract infection is an important factor associated with pin loosening during external fixation: a prospective...

The occurrence of pin loosening represents a common issue in the context of external fixation methodologies; nevertheless, a comprehensive investigation into the multifaceted causes of pin loosening, incorporating a multivariate analysis among pin infection, bone quality, and pin insertion angle, is notably absent in current literature. The present study endeavors to pinpoint factors associated with pin loosening through such a multivariate analysis.

The study encompassed patients who underwent the removal of external fixators from March 2023 to July 2023. The assessment of pin loosening was executed through the utilization of the pin track score, the pin removal torque value (PRTV), and the radiolucent zone around the pin (RZAP) as depicted in digital radiography (DR) images. Culturing of the pin-bone interfaces was performed, and measurements of the grayscale intensity of cortical bone (GSCB) and pin verticality within DR images were taken. Multivariate analyses were conducted employing a Generalized Linear Mixed-Effects Model (GLMM), Adjusted odds ratios (aORs) with 95% confidence intervals (CIs) were calculated by exponentiating the model coefficients (Exp(β)).

Altogether 47 patients with a total of 220 pins were included for analysis. The mean PRTV was 1.9 ± 2.1 N·m. The correlation analysis between PRTV and RZAP yielded a P-value of less than 0.001, signifying a substantial correlation between pin loosening and RZAP. For pins with a PRTV of 0, the RZAP measured 1.9 ± 0.8 mm. The positive rate of bacterial culture was 20%, and the loosening rate was 26.8%. Pin loosening was significantly associated with bacterial infection (aOR = 2.24, 95% CI: 1.03-4.90, P = 0.04) and GSCB (aOR = 0.50, 95% CI: 0.38-0.66, P < 0.01), but not with pin verticality (aOR = 1.00, 95% CI: 0.93-1.08, P = 0.99). Non-HA-coated pins remained significantly associated with bacterial infection (aOR = 8.20, 95% CI: 2.18-30.85, P = 0.002), whereas HA-coated pins were not (aOR = 3.44, 95% CI: 0.24-48.76, P = 0.36).

Pin loosening was significantly associated with bacterial infection at the pin-bone interface and lower GSCB, but not with pin verticality. Notably, infection strongly predicted loosening in non-HA-coated pins, while HA-coated pins demonstrated higher raw infection rates.